My final sketch is a squelchy electronic kinda thing, with an interesting bass sound playing an arpeggio with a clock-synced delay effect. The sound itself is something from the Blofeld’s wavetables, with an envelope used to modulate the wavetable entry, and some FM from a sawtooth wave to dirty it up a bit.
Another little classical piano kinda thing today, just arpeggiating my way through some chords. I like the progression, though, going down the scale from Am to E and coming back again, but with different variations along the way.
Today’s sketch is a bit of a rework of yesterday’s, tweaking the compressor settings on the vocals to even them up a bit more, and using volume automation to drop out a few of the loudest pops (those loud P and B sounds). I also re-did the extra piano part in the chorus at the end, since I wasn’t happy with yesterday’s. I could probably spend a whole day just tweaking the vocals to clean up the rough edges, but I’m happy enough with the results given the time I’ve spent so far.
Today, everyone I’ve beaten at SingStar gets a laugh at my expense 🙂 I’ve taken a stab at recording vocals, and the result — a cover of Four Seasons in One Day — gets a “not quite terrible” review from me. If nothing else, I got good practice working with Ardour, recording multiple tracks (two vocals, and two piano), splicing together bits of multiple takes, and using compression to iron out volume irregularities.
Another quick synth sound tonight — I played around with an analogue synth technique called hard sync, and came up with a nice cutting lead sound. It’s an interesting technique for making very bright sounds like this, so I’ll be keen to play with it more and see what other sounds I can make with it, especially using the Blofeld’s wavetables rather than just the simple sawtooth waves here.
I was pressed for time today, so I dived in to the Blofeld and pulled out a synth sound rather than doing anything terribly musical. Today I played with FM synthesis, which uses one audio signal to modulate the pitch of another, and I came up with a jazzy vibraphone/electric piano sound.
FM is a lot more complex and mathematical than analogue-style synthesis, but it’s an interesting tool, and it’s definitely well suited to metallic or bell-like sounds like this. The great thing about synthesising these sounds, instead of just using samples, is that you can sculpt the sound in realtime — this sound gets not just louder the harder you hit the keys, but brighter too, by increasing the amount of FM based on the key velocity.
I’m on the home stretch now — today’s the first day of the last week of this crazy experiment. Today’s sketch is an experiment in noise music in the most literal sense. I had some trouble with noise from the Blofeld while recording yesterday’s sketch. When I recorded that noise today I noticed some rhythmic elements in it, so I chopped it up and reassembled the bits in Ardour to produce a bit of a noisy percussion piece.
Of course, I usually don’t want to record this noise, but thankfully I’ve worked out how to kill it. It’s due to ground noise in the USB cable running from the PC to the Blofeld, so I just need to use the Blofeld’s standard MIDI input, rather than the USB connection.
I wasn’t musically inspired at all today, but I did think of a studio trick I’ve been meaning to try — double-tracking, where you record the same thing twice, panned hard left and right. It’s often used on vocals, because it gives them a nice wide sound.
No-one wants to hear me singing, so instead of vocals, I’ve tried it with the LinuxSampler piano, and with a simple synth lead/bass sound from the Blofeld. The sketch file has single-tracked and double-tracked versions of both, so you can hear the difference. I’m also trying a flash-based MP3 player widget thingy for my posts — let me know if it doesn’t work for you.
Today’s sketch is a bit of a throwback, both to my first sketch, and to my days of practicing for piano lessons when I was 16/17, when I’d get bored reading sheet music and start hammering out Beatles covers and random chords. Like the first sketch, it’s LinuxSampler’s great free piano providing the sound, with jconv providing a touch of reverb to open things out a bit.
The volume’s a bit rough — I need practice knowing just how hard to hit my CS2X keyboard — but apart from that I’m pretty happy with it; it feels like a chorus, so maybe I’ll have to come back to it next month to write some verses.
Another quick fiddle with the vocoder today. I wasn’t happy with how intelligible my speech was yesterday, so I recorded a vocal sample and then ran it through the vocoder using different carrier sounds from the Blofeld. The first is a simple sawtooth wave, same as I used yesterday, but played much lower on the keyboard; the second is my string sound, again played fairly low on the keyboard.
I could probably get better results with more fiddling, but both of these sound okay, and suitably robotic. I was having trouble with the vocoder, though — it’s fine when reading from a mic, but kept playing up when supplied with pre-recorded samples — so that’ll have to wait for another day.